BACKGROUND: Autism spectrum disorder (ASD) is a neurodevelopmental condition associated with sleep disturbances that may result from abnormalities in melatonin production. The correlations of melatonin levels with the severity of sleep disorder and/or severity of ASD were reported. OBJECTIVES: To evaluate urinary levels of the melatonin metabolite, 6-sulphatoxymelatonin (aMT6s), in children with ASD, and their associations with sleep abnormalities and behavioural impairments. METHODS: Study involved 77 children with ASD and 84 controls aged 2.5-15.5 years. Sleep disorders were assessed by Children's Sleep Habits Questionnaire. Morning and afternoon levels of aMT6s were determined by radioimmunoassay method. Urinary creatinine levels were assessed by an enzymatic method. RESULTS: The urinary aMT6s/creatinine values indicate that the night-time melatonin levels are signifi cantly lower in ASD than in controls, but there are no signifi cant differences in the daytime levels. In the ASD group, on average, a 6.8-fold difference between night-time and daytime values of urinary aMT6s/creatinine was found, whereas for the controls a 12.5-fold difference was observed, indicating a lower night-time increase in melatonin levels. In ASD group, the difference in night-time-daytime aMT6s/creatinine value correlated with some types of sleep problems, but not with the severity of ASD. CONCLUSION: The results indicate that in ASD there are differences in the patterns of melatonin secretion that may be associated with sleep impairment (Tab. 4, Fig. 2, Ref. 28).

In children with autism spectrum disorder (ASD), sleep disturbances are a frequent comorbidity with an adverse effect on their behavior and functioning. It was suggested that melatonin deficit is at least partly responsible for the sleep problems. The study aimed to investigate, in a sample of 56 children with ASD aged 2.8–13.3 years, if the sleep problems and melatonin secretion can serve as predictors of adaptive functioning and severity of the ASD core symptoms. We demonstrated that, after adjustment for age, the Sleep score assessed by the Children’s Sleep Habits Questionnaire predicts the Adaptive behavior composite score only in children younger than 6 years, and the preferred predictive model is for the domain Socialization. The age-adjusted Sleep score predicted Externalizing and Internalizing maladaptive behavior, with a near-zero contribution of age to the relationship between the Internalizing maladaptive behavior and Sleep score. After adjustment for age, the reduced night-time melatonin secretion predicted a higher severity of ASD symptoms in the domain Social affect and the Calibrated Severity Score, but not the sleep problems. Our results emphasize the importance of assessing sleep problems as a modifiable predictor of behavior in children with ASD and support the hypothesis about the role of melatonin in pathophysiology of ASD.